Preliminary clinical trials have demonstrated that radiolabeled anti-CD20 monoclonal antibodies can achieve remissions in 65-90% of lymphoma patients failing chemotherapy. However, most patients treated with conventional radiolabeled antibodies (RAb) subsequently relapse and die of recurrent lymphoma. The objective of this research proposal is to optimize radioimmunotherapy (RIT) of B cell lymphomas utilizing pretargeting amplification strategies to improve the efficacy and decrease the toxicity of conventional RIT. Two separate pretargeting approaches will be investigated, one using streptavidin (SA) and radioactive biotin and the second employing molecularly engineered bispecific anti-CD20 x anti-ligand antibodies which bind covalently to radiolabeled ligands. First, we will compare the biodistributions, toxicities and efficacies of 4 genetically engineered SA mutant molecules with native SA in combination with either biotin or a synthetic divalent bis-biotin targeting molecule in a mouse lymphoma xenograft model. These streptavidin mutants will afford a unique opportunity to test the effect of SA avidity on tumor penetration as delineated in the binding site barrier hypothesis. Next, we will compare the in vivo biodistribution, radiation dosimetry, and therapeutic efficacy of pretargeted &#946;-emitting radionuclides (90Y, 177Lu) with pretargeted &#945;-emitters (213Bi 211At, 225Ac) in murine xenograft models. Finally, we will evaluate the pharmacokinetics, biodistributions, toxicities and efficacies of novel molecularly designed bispecific anti-CD20 x anti-ligand Abs which possess a molecularly engineered binding pocket capable of binding covalently to synthetic radiolabeled electrophilic ligands. These bispecific anti-CD20 x anti-ligand Abs will be compared directly to the SA-biotin pretargeting approach in lymphoma models. We hypothesize that the pretargeting strategies defined in this proposal will improve the tumor-to-normal organ ratios of absorbed radiation compared with conventional RIT, allowing improvement in response rates and response durations with less toxicity than is currently feasible. We hypothesize that pretargeting will eliminate the necessity of administering myeloablative doses of 131I-anti-CD20 Ab with hematopoietic stem cell rescue to achieve maximal response rates and survival rates. We anticipate rapid translation of the results of these preclinical experiments into our clinical RIT program for human Non- Hodgkin's lymphomas.